Gazprom Neft: Digital instruments of increase in accuracy of well-drilling

2019: Development of the program

On April 1, 2019 the Scientific and technological center of Gazprom Neft announced development of the self-training program for cost optimization at construction of difficult horizontal wells. As noted in Gazprom Neft, replication of development on company assets will allow to reduce costs on creation of new wells approximately by 1 billion rubles and will reduce terms of their drilling. The innovative technology helps engineers of the company in real time by indirect parameters to specify geology of layer and, if necessary, to make the decision on correction of a trajectory of drilling.

At construction of wells it is necessary to monitor a drilling trajectory all the time to remain in borders of oil layer. For this purpose use the sensors transferring information on surrounding breed from drilling equipment. However even the most modern devices because of constructional features it is possible to locate only in 15-30 meters from a chisel. Data arrive with a certain delay, creating risk of an exit out of limits of a productive zone.

Gazprom Neft developed a digital instrument for the solution of this problem. The program uses machine learning quickly to analyze the parameters arriving directly from drilling equipment — a vibration level, speed of drilling and rotation of a rotor, load of a chisel. These indicators change depending on characteristics of layer, allowing to define quickly structure of breed, without waiting for receipt of data from sensors on the most boring tool. Such approach helps to manage quickly drilling in cases when the project arrangement of layer differs from actual, emphasized in Gazprom Neft.

According to the developer, the program itself studies during operation, the forecast of structure of surrounding breed becomes more precisely with each drilled meter. The technology can be adapted for work on different fields. It passed industrial tests on assets of Gazprom neft-Yamal — the accuracy of prediction of change of rock at well-drilling was 70%.